Introduction

Perioperative hidden blood loss (HBL) in orthopedic surgery represents a critical yet often under-recognized clinical challenge. Recent data indicate that 52.5%–82.2% of total blood loss in joint replacements and spinal procedures is attributable to HBL1,2,3. HBL exacerbates postoperative decrease in hemoglobin levels, resulting in increased transfusion requirements. In the absence of proper management, this condition can result in delayed wound healing, elevated risk of infection, and prolonged postoperative recovery times. As a serious complication in orthopedics, HBL is often overlooked clinically because of its covert nature. However, progressive anemia may even trigger severe outcomes, such as cardiovascular and cerebrovascular events, and cognitive dysfunction, directly affecting patient prognosis and quality of life4,5.

In geriatric populations, HBL-induced hemodynamic instability may contribute to increase fall-related mortality6,7. Despite advances in intraoperative hemostatic techniques, HBL episodes remain undetected until hemoglobin levels decline to a point that initiates transfusion protocols, highlighting systemic gaps in proactive monitoring. Several studies have investigated the dynamics of perioperative HBL in patients with intertrochanteric femur fractures3,4,8. Findings from these studies have helped identify early changes in hemoglobin levels, enabling prompt intervention and improved outcomes.

The concept of HBL was first introduced in 200099. Consequently, many nurses have not received adequate education on the assessment and management of HBL. Although a standardized prevention strategy for perioperative HBL among orthopedic patients is still lacking, early evaluation, identification, and intervention can mitigate its adverse effects on surgical outcomes10. Most contemporary studies have focused primarily on elucidating the pathophysiological mechanisms of HBL11,12. However, emphasis on monitoring HBL remains inadequate. Nonetheless, monitoring constitutes the primary line of defense against HBL complications. Nurses are encouraged to adopt a multifaceted monitoring approach that integrates physiological indicators (e.g., hemodynamic trends), biomarker analysis (e.g., hemoglobin drift patterns), and clinical symptom recognition (e.g., fatigue, dizziness, and shortness of breath). The incorporation of these elements enables the early identification of patients with HBL, thereby facilitating predictive intervention and averting adverse clinical outcomes attributable to HBL13,14,15.

The Knowledge-Attitude-Practice Model (KAP Model) posits that the professional knowledge base of healthcare personnel exerts a direct influence on clinical decision-making behaviors, which, in turn, are influenced by cognitive attitudes16. Analysis of the current clinical context and qualitative interviews with nurses revealed several challenges in HBL management, including limited awareness, inadequate knowledge proficiency, non-standardized assessment procedures, and fragmented intervention strategies. Fundamental knowledge, positive attitude, and proper practices are essential prerequisites for enhancing the quality and safety of nursing care17,18. Nurses’ risk perception directly influences the implementation of preventive measures; those who can identify patients at a high risk of HBL can intervene promptly. Therefore, understanding the knowledge, attitudes, and practices (KAP) of orthopedic nurses toward the assessment and management of perioperative HBL will enable the implementation of targeted training programs to bridge the gap between knowledge and clinical practice. However, limited studies have systematically mapped the KAP dimensions of the assessment and management of HBL among orthopedic nurses, thereby leaving a gap in the evidence necessary to train nurses on HBL management. This multicenter, cross-sectional study aimed to:

  1. (1)

    Quantify the current levels of KAP among orthopedic nurses regarding the assessment and management of perioperative HBL.

  2. (2)

    Identify modifiable factors influencing KAP performance, including age, work experience, hospital level, professional title, highest educational attainment, and whether the nurses had received orthopedic specialist training.

  3. (3)

    Establish a theoretical foundation for developing an evidence-based educational framework to improve the assessment and management of perioperative HBL, thereby contributing to enhanced patient safety in clinical practice.

Methods

Design

This was a descriptive, multicenter, cross-sectional survey. A schematic overview of the study design is shown in Supplementary Figure S1.

Participants and settings

Between June 1 and September 30, 2023, convenience sampling was used to survey orthopedic nurses from 40 hospitals across 10 Chinese provinces. To ensure representativeness and generalizability of the results, sample selection was based on a range of hospital levels and patient distribution. Approximately 80% and 20% of the participants were selected from tertiary and secondary hospitals, respectively. This study was approved by Zhengzhou University Life Science Ethics Review Committee (No. 2022-03-18). All methods were performed in accordance with the relevant guidelines and regulations, and in accordance with the Declaration of Helsinki. The study adhered to the STROBE checklist for cross-sectional studies19 (see Document S1). According to the Kendall method for sample size calculation, the required sample size was calculated to be 5–10 times the number of variables20. The sample size was increased by 20% to account for potential sample loss. Therefore, the calculated sample size was between 336 and 672, and was considered appropriate for this study. The study team contacted the head of nurses in the orthopedic departments of each hospital. After obtaining permission, the team shared the survey link on WeChat. The survey included an explanation of the study’s purpose and significance, and informed consent was obtained from orthopedic nurses who participated voluntarily. An electronic questionnaire required all questions to be answered. Incomplete or missing responses prevented submission, and only one submission per device was allowed. Participants whose responses displayed clear patterns or identical answers across all items were excluded.

Inclusion criteria for the study participants were a minimum of 1 year of work experience in orthopedics, possession of a nursing license, and employment at the time of the study. Nursing interns, trainees, and nursing assistants were excluded. The study was conducted as part of a departmental quality improvement initiative and registered with the hospital.

Data collection instruments

The survey questionnaire was divided into three parts: demographic information; items assessing nurses’ KAP regarding the assessment and management of perioperative HBL in orthopedic patients; and three open-ended questions.

Demographic information

Demographic information about the nurses, including age, work experience, hospital level, professional title, highest educational attainment, and whether they had participated in specialized training in orthopedic nursing, was collected.

KAP assessment scale

A questionnaire titled “Knowledge, Attitudes, and Practices of Orthopedic Nurses in the Assessment and Management of Perioperative HBL in Orthopedic Patients” was developed based on literature reviews8,21,22. The 56-item questionnaire comprised three parts as follows:

  1. (1)

    Knowledge Dimension: Comprise concepts, incidence, clinical manifestations, adverse outcomes, predisposing and triggering factors, diagnosis, and treatment of perioperative HBL (32 items). Knowledge levels were rated on a scale from “completely unaware” (1 point) to “fully mastered” (5 points), with a total score between 32 and 160. Higher scores indicated greater mastery of knowledge regarding the assessment and management of perioperative HBL.

  2. (2)

    Attitudes Dimension: Nurses’ willingness to learn about HBL and their attitudes toward assessing and managing perioperative HBL, as well as the barriers affecting their attention to HBL assessment and management, was assessed in this dimension; this comprised 13 items. Barriers to assessment and management were negatively framed and scored in reverse order. Response options ranged from ‘strongly disagree’ (1 point) to ‘strongly agree’ (5 points), with total scores between 13 and 65. Higher scores indicated a more positive attitude toward learning about HBL assessment and management.

  3. (3)

    Practice Dimension: Willingness to engage in clinical practices regarding the assessment and management of perioperative HBL was assessed in this dimension; this comprised 11 items. The response options ranged from “never” (1 point) to “always” (5 points), with a total score between 11 and 55.

After developing an initial draft of the questionnaire, it was transformed into an evaluation form and reviewed for content validity by eight orthopedic medical experts (all chief physicians), eight orthopedic nursing managers (two chief superintendent nurses and six co-chief superintendent nurses), and four nurse practitioners (deputy chief nurses). The item-level content validity index was between 0.70 and 1.00, and scale-level content validity index was 0.95. Prior to the multicenter survey, a pilot study was conducted with 100 orthopedic nurses (a single-center convenience sample) from the authors’ institution to assess item clarity and preliminary reliability. Importantly, the 100 pilot responses were used solely for instrument refinement and were not included in the final analytical dataset of valid survey responses. Cronbach’s alpha coefficients demonstrated strong internal consistency, with an overall α of 0.859, and subscale values of 0.908 (knowledge), 0.792 (attitudes), and 0.873 (practices). These results indicate good-to-excellent reliability across the subscales, confirming the suitability of the instrument for use in a multicenter study.

Open-ended questions

This study also included three open-ended questions, which are as follows:

  1. (1)

    How can orthopedic nurses learn to assess and manage perioperative HBL?

  2. (2)

    What specific contents are orthopedic nurses most eager to learn regarding the assessment and management of perioperative HBL?

  3. (3)

    What barriers do orthopedic nurses encounter when implementing perioperative HBL assessment and management?

Data analysis

Data were organized and analyzed using Excel and SPSS (version 27.0; IBM Corp., Armonk, NY, USA). Categorical data are described as frequencies and percentages. Continuous variables with a normal distribution are expressed as mean and standard deviation (SD). Scoring rate was calculated by expressing the actual score for each dimension as a percentage of the total possible score. Normality of continuous variables was evaluated using Q–Q plots as well as skewness and kurtosis statistics. Given the large sample size (N = 456), these distributional indices were considered more informative than formal tests of normality, which are highly sensitive to trivial deviations in large samples. The Pearson’s correlation coefficient was used to examine the relationships between KAP. Independent t-tests were used for comparisons between two groups, whereas analysis of variance (ANOVA) was used for comparisons among multiple groups. Multiple linear regression analyses were conducted to identify the factors influencing nurses’ KAP toward perioperative HBL. For univariate analyses, effect sizes were reported as the Cohen’s d for two-group comparisons and η² for ANOVA. For non-parametric tests, r (|Z|/\(\:\sqrt{N}\)) or η²_H for Kruskal–Wallis was calculated. Effect size thresholds were defined as follows: 0.01 ≤ η² < 0.06 indicates a small effect, 0.06 ≤ η² < 0.14 indicates a medium effect, and η² ≥ 0.14 indicates a large effect23. According to the Cohen’s criteria for interpreting d in t-tests, a value of 0.2 represents a small effect, 0.5 a medium effect, and 0.8 a large effect24. For regression models, the R² (adjusted R²) and Cohen’s f² (overall and incremental) values were reported. The thresholds for small, medium, and large effects followed the Cohen’s conventions (0.02 = small, 0.15 = medium, and 0.35 = large)25. Statistical significance was set at p < 0.05.

Results

Demographic characteristics

A cross-sectional survey was conducted across 40 hospitals located in diverse regions of China, including the Central, Northern, and Southern provinces, as well as the plateau, inland, and coastal areas. Of 480 nurses invited to participate in the study, 24 were excluded owing to invalid responses. Ultimately, 456 nurses participated in the study, yielding an effective response rate of 95.00%. Mean age of the 456 participating nurses was 33.45 (7.03) years and all were female. The patients’ basic characteristics are presented in Table 1.

Table 1 General demographic characteristics of participants (N = 456).
Full size table

Normality assessment

Normality of continuous variables was evaluated using Q–Q plots as well as skewness and kurtosis statistics. Skewness values ranged from − 0.371 to − 0.978 and kurtosis values ranged from − 0.030 to 0.599, all well within widely accepted thresholds for approximate normality (|skewness| < 2, |kurtosis| < 7)26. Although the Q–Q plot for the knowledge score exhibited a slight deviation in the lower tail, the overall distribution remained close to normal. Given the large sample size (N = 456) and the near-symmetric distributional characteristics, parametric analyses were deemed appropriate. The corresponding results are presented in Figs. 1, 2 and 3; Table 2. Additionally, Likert-scale total scores were treated as approximate interval data, which is a widely accepted approach in large-sample nursing studies, further supporting the appropriateness of parametric testing in this study.

Figs. 1
Figs. 1
Full size image

Q–Q Plot of knowledge total score.

Figs. 2
Figs. 2
Full size image

Q–Q Plot of attitude total score.

Figs. 3
Figs. 3
Full size image

Q–Q Plot of practice total score.

Table 2 Descriptive statistics with skewness and kurtosis values.
Full size table

KAP status on the assessment and management of perioperative HBL among orthopedic nurses

Mean knowledge score for the assessment and management of perioperative HBL among orthopedic nurses was 120.16 (SD = 30.85), with a scoring rate of 75.10%. Mean attitude score was 50.63 (SD = 9.59), corresponding to a scoring rate of 77.89%. Mean practice score was 40.17 (SD = 9.33), with a scoring rate of 73.03%. Table 3 summarizes the top and bottom three items for each dimension (knowledge, attitude, and practice). Results of the open-ended questions are presented in Table 4.

Table 3 Top and bottom three scores of the knowledge, attitude, and practice of nurses toward perioperative HBL assessment and management.
Full size table
Table 4 Results of open-ended questions.
Full size table

Correlation between KAP regarding the assessment and management of perioperative HBL in orthopedic nurses

Table 5 presents the correlations among orthopedic nurses’ knowledge, attitudes, and practices regarding perioperative HBL. Knowledge was positively correlated with practice (r = 0.608, p < 0.05), indicating that nurses with higher levels of knowledge tended to demonstrate better HBL-related clinical practices. Attitude showed a weak negative correlation with practice (r = − 0.118, p < 0.01), indicating that higher attitude scores were associated with slightly lower practice scores. No significant correlation was observed between knowledge and attitude (r = − 0.075, p > 0.05).

Table 5 Correlation between knowledge, attitudes, and practices in related to the management of hidden blood loss by orthopedic nurses.
Full size table

Associations between demographic characteristics and KAP toward perioperative HBL

Results of the univariate analyses and corresponding effect sizes are presented in Table 6. Knowledge and practice scores were significantly associated with age, work experience, hospital level, professional title, highest educational attainment, and participation in specialized training in orthopedic nursing (p < 0.05). Attitude scores were significantly associated with work experience, professional title, highest educational attainment, and participation in specialized training in orthopedic nursing (p < 0.05). The effect size estimates indicated that most associations were of small to medium magnitude, whereas the hospital level demonstrated a large effect on knowledge and medium effect on practice (Table 6).

Table 6 Demographic characteristic correlations with knowledge, attitudes, and practices in the assessment and management of perioperative HBL in orthopedic patients (N = 456).
Full size table

Multiple linear regression analyses of KAP in relation to demographic characteristics

A multiple linear regression model was used to identify predictors of KAP (Table 7). Age, work experience, hospital level, professional title, highest educational attainment, and whether the participants had received specialized training in orthopedic nursing were treated as categorical variables and dummy coded. For knowledge, significant predictors included work experience, hospital level, professional title, highest educational attainment, and participation in specialized training in orthopedic nursing (p < 0.05). For attitude, significant predictors were age, work experience, hospital level, highest educational attainment, and participation in specialized training in orthopedic nursing (p < 0.05). For practice, hospital level, educational attainment, and participation in specialized training in orthopedic nursing were significant predictors (p < 0.05). No evidence of multicollinearity was detected in any regression model. All the models demonstrated large effect sizes (f² ≥ 0.35), indicating strong explanatory power of the included predictors (Table 8).

Table 7 Multiple linear regression analyses of knowledge, attitudes, and practices scores for orthopedic nurses with demographic characteristics.
Full size table
Table 8 Magnitude of effect in regression.
Full size table

Discussion

HBL is a frequent and often underestimated complication of orthopedic procedures such as hip fracture surgery27, total knee arthroplasty9, hip arthroplasty28, and spinal surgery3. It is a major cause of postoperative anemia, which increases the risk of infection and mortality, prolongs hospitalization, and impairs functional recovery and quality of life. Therefore, effective assessment and management of HBL are critical components of perioperative blood management. Previous studies have indicated that adequate nursing knowledge is essential for effective clinical decision-making and patient management29. In this study, orthopedic nurses demonstrated moderate knowledge on perioperative HBL, underscoring the need for targeted education and training. Despite general positive attitudes, low practice scores revealed a gap between intention and implementation, indicating that theoretical understanding has not yet translated into consistent clinical performance. Enhancing procedural competence and establishing standardized protocols among nurses are key to improving management of perioperative HBL.

Factors affecting knowledge toward perioperative HBL

The mean knowledge score rate of orthopedic nurses toward the assessment and management of perioperative HBL was 75.10%, indicating a moderate level of understanding. Knowledge was significantly influenced by work experience, hospital level, professional title, educational attainment, and participation in specialized training in orthopedic nursing (p < 0.05).

Consistent with previous studies, professional competence increased with clinical exposure and institutional support because nurses with more experience, higher professional titles, or positions in tertiary hospitals have greater access to knowledge and continuous education30,31,32,33,34. Higher educational attainment and professional certification, particularly in orthopedic nursing, enhance nurses’ technical competence, confidence, and ability to integrate new evidence into practice35,36. This aligns with our finding of a positive correlation between knowledge and practice (r = 0.608, p < 0.05).

As medical technology continues to advance, continuous education and training remain key determinants of effective HBL management. Hospitals should strengthen investments in educational initiatives and specialist training, encourage professional certification, and promote multidisciplinary learning environments to reinforce evidence-based clinical decision-making.

Factors affecting attitude towards perioperative HBL

Nurses achieved the highest mean attitude score (77.89%), reflecting a generally positive outlook towards perioperative HBL management and strong willingness to engage in related clinical activities. Attitude was significantly influenced by work experience, professional title, educational level, and participation in specialist nursing training. Consistent with previous studies, nurses with greater clinical experience, higher professional titles, or advanced education demonstrated stronger competence and confidence, resulting in more positive professional attitudes37,38. Certification as an orthopedic specialist nurse further reinforces professional commitment and confidence39.

However, the weak negative correlation between attitude and practice (r = − 0.118, p < 0.01) suggests that favourable attitudes do not necessarily lead to consistent behaviour. Social desirability bias and contextual barriers, such as time constraints, staff shortages and competing workloads, may prevent positive attitudes from being expressed in daily practice. Reverse-scored items, partial overlap between cognitive attitudes and behavioural practices, and response-style tendencies in online surveys may introduce measurement artefacts that attenuate or invert the expected association. Together with contextual constraints, these methodological factors explain the weak negative correlation observed.

Factors affecting the practice of perioperative HBL assessment

The mean practice score (73.03%) was the lowest among the three dimensions, indicating limited hands-on application. In China, HBL assessment remains predominantly physician-led and relies on complex formulas (e.g., Gross and Nadler) and laboratory indices that restrict nurses’ direct participation40. Additionally, ward nurses carry a heavy clinical workload that encompasses fundamental care, complication monitoring, and coordination of multidisciplinary activities, including postoperative vital sign surveillance, prevention of deep vein thrombosis, wound management, and organization of consultation. In tertiary orthopedic wards, low nurse-to-patient ratios and limited staff further constrain nurses’ involvement in HBL assessment and intervention41.

Practice was significantly influenced by work experience, hospital level, educational attainment, and specialist training (p < 0.05). This finding is consistent with that in a previous study by Carless-Kane et al., who reported that nurses with higher educational levels were more capable of applying theoretical knowledge to practice, particularly in complex clinical situations such as postoperative HBL42. Similarly, specialist certification was associated with greater engagement in assessment and stronger competence in decision-making.

The hidden nature of HBL often results in a lack of obvious clinical symptoms and reduces its detectability in routine nursing assessments. However, advances in point-of-care hemoglobin analyzers and digital monitoring systems now enable more active nursing participation43,44. Collaborative practice further enhances early detection and timely intervention. For example, Mousa et al. demonstrated that certified nurses within multidisciplinary teams identified HBL within 24 h postoperatively, significantly improving intervention efficiency45.

Although attitudes were positive, structural barriers continued to hinder consistent implementation, emphasising the need for enhanced training, staffing and interprofessional collaboration. The discrepancy between attitudes and practice may also be due to measurement factors, such as scale structure and response styles, which limit the correspondence between self-reported attitudes and actual behaviours. Future work should examine these methodological factors in more detail.

Magnitude of effects and implications

Effect size analyses further clarified the relative influence of demographic and professional factors on nurses’ perioperative HBL KAP. In the univariate analyses, most variables (e.g., age, work experience, and educational attainment) showed small effects (η² = 0.014–0.051; Cohen’s d ≈ 0.2–0.3), whereas hospital level had a large effect on knowledge (η² = 0.218) and medium effect on practice (η² = 0.096), underscoring the importance of institutional context. Specialist orthopedic nurse training produced medium effects across all three dimensions (Cohen’s d = 0.521–0.762), while professional title had moderate effects on knowledge and attitude but small effects on practice. The multivariate regression models yielded large overall effect sizes (f² > 0.35), confirming that the combined predictors explained a substantial proportion of the variance in KAP outcomes. These findings emphasize the need to strengthen individual competencies and institutional support systems.

From institutional strategies to policy-level integration

Persistent discrepancies between nurses’ attitudes and practices highlight the need for workforce-level interventions, including tiered training systems, continuous professional development, and broader involvement in multidisciplinary decision-making. To improve consistency in clinical practice, hospitals should optimize staffing levels, standardize HBL assessment procedures, and provide clearer operational guidance. Strengthening institutional support will help nurses translate their knowledge and positive attitudes into routine clinical behaviors. At the policy level, incorporating HBL management into national nursing standards, linking it with quality improvement initiatives, and promoting balanced interprofessional collaboration would further enhance implementation across institutions.

Conclusion

The KAP of orthopedic nurses toward perioperative HBL remain moderate, underscoring the urgent need for targeted strategies to enhance clinical competence. Strengthening structured training, promoting continuous professional development and specialist certification, and integrating nurses into multidisciplinary decision-making are key priorities. These measures will help align individual competencies with institutional expectations, enabling nurses to translate knowledge into practice, thereby improving the quality and safety of orthopedic nursing care. Beyond the local context, our findings provide insights into global nursing practice by highlighting the importance of workforce capacity building to address HBL and other perioperative risks, ultimately contributing to health equity worldwide.

Limitations

This study has several limitations. First, all data were collected via self-reported questionnaires, which may have introduced social desirability and common method biases. Second, recruitment through head nurses and the online survey format may have led to a sampling bias. Third, all participants in this study were female, reflecting the gender distribution typical of orthopedic nursing in China. However, this homogeneity limits the generalizability of the findings, as potential gender-related differences in knowledge acquisition, clinical decision-making, and professional attitudes could not be examined. As a result, the KAP patterns identified in this study may not fully represent male nurses or more gender-balanced nursing teams. Future studies should aim to include more diverse samples to determine whether sex-related factors influence nurses’ assessment and management of perioperative HBL. Fourth, minor deviation was observed in the lower tail of the knowledge score distribution, but the overall skewness and kurtosis values supported approximate normality. Given the large sample size and the robustness of parametric tests under mild non-normality, this analytical approach was deemed appropriate; however, the results should still be interpreted with caution.

Although this study selected representative hospitals from different regions in China, it did not include hospitals from all regions of the country. Therefore, the findings may not accurately reflect the current KAP state of orthopedic nurses regarding nationwide assessment and management of HBL. Nevertheless, to our knowledge, this is the first multicenter study to focus specifically on the KAP of orthopedic nurses regarding perioperative HBL assessment and management. Future studies should expand the geographic coverage using stratified sampling, increase the sample size, perform subgroup analyses to enhance generalizability, and achieve a more comprehensive understanding of HBL management practices.